Honey, Hydrophobicity and Biofilms

A reader (Jay Bryant) recently pointed out a PNAS article on the structure of a bacterial enzyme that uses sucrose to make the glucan matrix of dental biofilms.  This article released a cascade of associations in my mind and illustrated why honey does not contribute to dental plaques, but is antimicrobial and aids wound healing.  People forget that sugars combine both hydrophilic and hydrophobic properties, and thereby act as soaps.

The starting point of the chemical versatility of carbohydrates is the inability of the central portion of a sugar ring structure to form hydrogen bonds.  Each sugar is made of  a linear chain of carbon atoms with each carbon linked also to a hydrogen and a hydroxyl.  Only the hydroxyl can participate in hydrogen bonds, so each carbon has a hydrophilic side (bonds with water to make hydrogen bonds) and a hydrophobic side (that makes van der Waals bonds with other hydrophobic molecules.)  The sugars circularize and the rings have faces that are predominantly hydrophobic and perimeters with hydroxyls that are hydrophilic.  Polysaccharides (long chains of sugars), such as cellulose, can sometimes form long fibers that form a hydrophobic context for hydrogen bonds between the hydroxyls of adjacent polymers.  These cellulose fibers are very resistant to chemical or biological attack and accumulate as the most abundant biological molecules on Earth.

The PNAS article provides another example of how protein enzymes interact with carbohydrates, in this case sucrose and a polymer of glucose.  Typical weak bonds between the amino acid residues of proteins and other molecules are hydrogen, ionic or van der Waals bonds with energies of a couple of kcals/mol.  In contrast, the bonding of the hydrophobic face of a sugar to the hydrophobic face of an hydrophobic amino acid, e.g. tryptophan, phenylalanine, histidine, lysine or arginine, releases more than ten kcals/mol of energy.  Thus, the structure of the bacterial enzyme that makes biofilm glucan chains from dietary sucrose, the sucrose is bound to the enzyme on the face of a prominent tryptophan.  Examination of enzymes that bind to polysaccharides will show a series of tryptophans arrayed across the surface of the enzymes with spacing appropriate to bind to the individual sugars of the polysaccharide.

Biofilms are communities of multiple species of bacteria held together by a polysaccharide matrix.  In the case of dental plaque, the polysaccharide is made of glucose links, whereas many other matrix polysaccharides are negatively charged and held together by positively charged metal ions.  The bacteria bind to the polysaccharides using protein receptors that exploit the display of hydrophobic binding sites of the polysaccharides.  It takes energy to make polysaccharides and the dental plaque bacteria use the energy already expended in the formation of sucrose to produce a polymer of glucose, an alpha-glucan, and free fructose.  Thus, sucrose is essential in forming this type of biofilm and without this sugar, the dental plaque cannot form.  Milk lactose, or glucose would be a more appropriate sweetener.  Unfortunately, high fructose corn syrup would be a poor substitute, because of the high liver toxicity of the fructose (it causes fatty liver, just like alcohol) and very high activity in forming advanced glycation end products (AGEs), which contribute to the symptoms of  diabetics.

Honey seems to be magical, because at low concentrations the sugars present in honey  (mostly glucose and fructose, and not sucrose) are nutrients for bacteria, but at high concentrations honey is anti-bacterial and useful as a wound treatment.  I think that the explanation for its antimicrobial activity is that sugars are amphipathic, that is they have both hydrophilic and hydrophobic properties, just like soap, and at high concentrations they kill bacteria, just as soaps at high concentrations kill bacteria.  In fact, the gentle soaplike properties of sugars are exploited experimentally to dissolve proteins that are normally embedded in cellular membranes.  This explanation predicts that corn syrup, which can also be used to form very stable soap bubbles, should also be useful in wound healing.

Lectins - Heat’em and Eat’em

Lectins are proteins common in seeds. They bind to sugars attached in chains to proteins, i.e. glycoproteins, and are displayed on the surfaces of cells that line the gut. Lectins could inhibit digestion of raw beans, but cooking makes them digestible.

Fear of lectins is puzzling. Lectins are proteins that have binding sites on their surfaces for specific single or small sequences of sugars. They are present in seeds to protect the seeds from herbivores.

A seed is mostly food (starch, protein, fat) for the plant embryo that will grow from it. This is also true of a chicken egg and just like the egg, the seed contains defensive proteins to inhibit the growth of bacteria, fungi and egg/seed eaters.

The egg has enzymes to degrade bacterial walls and proteins that bind iron, vitamins, etc. needed by bacteria and humans. Eating many raw eggs can lead to vitamin deficiencies. Boiling the eggs, unravels the defensive proteins and makes them digestible and nutritious.

Seeds block being digested by containing proteins that foul the digestion system of would be devourers. For example, soybeans have trypsin inhibitor that binds to our digestive enzyme and makes eating raw soybeans nonproductive and uncomfortable. Boiling soybean meal to produce a curd, i.e. tofu, agglutinates the denatured soy proteins, including the lectins and washes away the soy trypsin inhibitor. Tofu is free of digestion inhibitors and lectin activity.

It is not an accident that lectins bind to human red blood cells. The sugars displayed on the surface of red blood cells are the blood group antigens. Different sugars on the end of the sugar chains decorating RBCs characterize the A, B and O antigens. These same sugars are present on the surfaces of various bacteria. Immune systems don’t produce antibodies to self antigens, so a person with type A blood produces antibodies only to B antigen sugars it encounters on bacteria. A person who is type AB doesn’t produce antibodies to A or B sugar antigens. There aren’t antibodies to O, because that sugar structure is the basis upon which both A and B are made, and some of the
structure is present on all RBCs. Lectins are specific for A or B or other common bacterial sugar antigens.

I did some modeling to show a lectin with lactose (red and gray) bound to sticky tryptophans (yellow) in two places on the surface. In one case a lysine (blue) is draped on the other side. That shows that sugars bind both to aromatic amino acids and to the hydrophobic arms of basic amino acids.


Some people think that humans and other mammals must be protected from lectins and that this protection is shown in human and cow’s milk in the form of antibodies against lectins. This seems to be a misunderstanding. For example, human antibodies secreted in breast milk are secretory IgAs. These antibodies are glycoproteins, i.e. they are proteins with attached sugar chains. Some lectins will bind to these antibodies, because of the attached sugars. These are not lectin-specific antibodies, but rather glyco-specific lectins. The lectins are binding to the glycoprotein antibodies, not the other way around.

It is possible for people to be allergic to lectins, but this is unlikely. For example, peanut allergies involve proteins other than the peanut lectins.

There are some dangerous lectins. For example, ricin is a very nasty, but effective, toxin produced by the castor plant. Ricin is a lectin, in that it binds very specifically to sugars found on the surface of gut cells of insects and humans. After the ricin binds to surface proteins, it is brought into the cells where it chops up the protein synthesizing machinery. That is a dangerous lectin. It takes very little ricin to kill each cell and only a tiny amount to kill a human. Ricin is a terrorist toxin. Yet oil extracted from castor beans contains so little contaminating ricin that it is safe to eat. [Castor oil is wonderful to apply to aching feet overnight for painfree, luxuriously soft feet in the morning.]

The bottom line is that seed lectins add to the nutrition of cooked beans and grains that have been the foundations for several thriving civilizations. The longest living members of the bean and grain cultures are typically older and more fit than comparable individuals with a modern, inflammatory diet based on omega-6 oils.

Turkey, Tryptophan & Transmitters

Does eating turkey make you stuffed and drowsy? It must the be tryptophan... or not. Sure there is tryptophan in turkey and tryptophan is the starting point for making some neurotransmitters and hormones, but turkey meat is simply protein and fat. Tryptophan (left) is just one of the twenty amino acids found in most proteins, so the drowsiness after the big Turkeyday dinner is more about the “big” and less about the turkey.

Turkey meat is muscle and muscle, as the diagram shows is made of protein molecules that use energy in the form of ATP to move past each other and contract. We chew up the turkey muscle and our stomach juices contain enzymes (these “proteases” are also proteins) that reverse the process of protein synthesis and produce protein fragments called peptides. The specificity of the proteases in the stomach, e.g. pepsin, results in peptides containing intact heparin-binding domains that are also antimicrobial. In the intestines, a new group of proteases are added by the pancreas and the peptides are further reduced in size and heparin-binding domains are degraded. [Pathogens need heparin-binding domains to bind to the intestines.] The remaining peptides bind to the microvilli of the endocytes lining the small intestines, surface bound peptidases release individual amino acids and transport proteins bring amino acids into the endocytes and on to the blood stream.

In the brain, tryptophan is converted by a series of enzymes into serotonin and the serotonin is stored in secretory vesicles adjacent to the synapse that controls signals between nerves. A nerve action potential moves down the axon from the cell body to the synapse. The change in electrical potential reaching the synapse causes the secretory vesicles to fuse with the cytoplasmic membrane and release the serotonin into the synapse. The serotonin binds to the receptors of the adjacent nerve and starts a new action potential that travels to the next nerve body to repeat the process. The synapse is reset by reuptake or degradation of the serotonin. The degradation product, 5-HIAA, is removed into the blood and excreted in urine.

Turkey tryptophan does get converted into serotonin and high serotonin could make you mellow, but turkey is just like any other meat source of tryptophan. The big meal just makes you groggy, because there is less blood to your brain when the mysenteric blood flow is enhanced for digestion. There may also be a rise in blood sugar as the starch off your plate is rapidly converted into glucose in your blood. The potentially damaging high blood sugar is controlled by a rise in insulin that lowers glucose in the blood by stimulating transport into fat cells for immediate conversion into fat. The starch from the meal is rapidly depleted, blood sugar rises and then sudden falls. The low blood sugar also leaves you groggy.

So it was the size of the meal (decreased brain blood flow) and the sweet potatoes and rolls (starch-induced hypoglycemia) that induced you to kickback on the sofa and pass out with the big game lulling you to sleep. Tryptophan from the big bird is in the background waiting for you to awaken. Before you take the first mouthful, check to make sure that your meal follows the anti-inflammation guidelines. Planning ahead can help you to enjoy a meal that won't be a pain later.

Topoisomerase Inhibitors

Inhibiting enzymes involved in DNA synthesis should stop cancer cells, because cancer is uncontrolled cell division. Topoisomerases are enzymes that help to relieve the twists on double helical DNA as it unwinds preparatory to replication. It appears logical that topoisomerase inhibitors should be cancer inhibitors. Unfortunately targeting DNA-binding proteins also targets most of the signal receptors that are the targets for the evolution of plant alkaloids.

Drugs are designed to be specific in their interactions with a particular target protein, but they are too small to be specific and end up binding to many other related proteins. Hence, drugs have side reactions that are to some extent unpredictable, because the interacting proteins are not known.

Aspirin, for example, is supposed to bind to and inhibit COX-2, the enzyme that converts omega-3 and omega-6, long-chain fatty acids into corresponding anti-inflammatory and inflammatory prostaglandins, resp. Aspirin also binds to proteins that inhibit NFkB, the transcription factor that controls expression of inflammatory genes. Aspirin binds to dozens of other proteins. Aspirin does lots of other things than just blunt inflammation, but those side reactions are usually not significant enough to get our attention.

Heparin is one of the most commonly used drugs. It binds to and activates an inhibitor of thrombin, an enzyme that activates fibrin and mediates clotting. Heparin also binds to other components of the clotting system, as well as a dozen components of the complement system, and most of the cytokines that control communications throughout the body. When patients are given heparin injections, heparin binds continually to all of these components and must be constantly supplemented and monitored. Inflammation depletes the heparin components throughout the body, so it is not known prior to injection, how much heparin will be needed to saturate other serum proteins before the desired level of clotting inhibition is achieved. This illustrates rather dramatically that most drugs have only limited specificity.

One of my students provided another example of the minimal specificity of small molecules, especially the alkaloids and phenolics produced by plants. He brought to me a research article espousing the use of phenolics from yerba mate, which serves as a coffee-like stimulant in Argentina, as a topoisomerase inhibitor and potential anti-tumor treatment. Sure enough, phenolics extracted from this plant inhibit topoisomerase, and they may well be able to inhibit the growth of tumors, but it is doubtful that the binding of the phenolics to topoisomerase in the tumor nuclei has anything to do with inhibition of tumor growth.

Topoisomerase binds to nuclear DNA as the DNA unwinds during replication to produce two new double helical DNA molecules. Topoisomerase is a DNA-binding protein, i.e. a protein that binds to a negatively charged polymer of small deoxyribose sugars and flat purine and pyrimidine bases. Proteins bind to DNA in two ways. Amino acids of the protein either bind along the edges of the hydrophobic stack of base pairs, e.g. sequence-specific transcription factors, or they provide hydrophobic, flat surfaces that bind to the hydrophobic faces of the separated bases. Topoisomerase does both, because it deals with single-stranded regions of DNA and therefore binds to both the phosphates, as well as the bases. The important point here is that both aromatic amino acids, with flat hydrophobic rings, and the hydrophobic tails of basic amino acids, i.e. lysine and arginine, bind to the hydrophobic faces of nucleic acid bases.

I have illustrated the binding of a “topoisomerase inhibitor” to show the arginine (blue) in the active site cleft of the topoisomerase that binds across the hydrophobic face of the inhibitor (grey and red). Many plant phenolics and alkaloids would be expected to similarly bind and act as inhibitors of topoisomerase. This observation and the ease by which alkaloids enter cells (attached to circulating heparan sulfate?) suggests that a major function of the nuclear envelope may be to minimize access of alkaloid and related molecules to the nucleic acid binding proteins of the nucleus.

The binding promiscuity of secondary plant products is further exemplified by berberine. Berberine is an alkaloid found in goldenseal and is an herbal remedy used to treat a variety of inflammatory diseases. It also binds to heparin (and nucleic acids) to produce a fluorescent complex. Thus, mast cells that store and secrete histamine and heparin to produce the symptoms of allergy, can be vividly stained with berberine.

I could not resist the temptation to check to see if berberine also binds to topoisomerase. A quick search of the research literature showed that berberine is in fact a topoisomerase inhibitor.

The numerous cross reactions of drugs are further illustrated by metformin, the common drug used in the treatment of type II diabetes. Metformin is approximately planar and provides a surface that cannot hydrogen bond, i.e. it is hydrophobic. I expected that metformin would bind to tryptophans that I observed as common substrate-binding amino acids in the active sites of proteins that bound to polysaccharides, e.g. lectins, glycosidases and glycanases. To test this, I had students in one of my courses examine the inhibitory activity of metformin on E. coli beta-galactosidase. They found measurable inhibition and support for competitive binding to the active site that contains a pair of the predicted tryptophans.

My protein modeling and structural studies show the basis for numerous interactions between plant secondary compounds, drugs, nucleic acids, polysaccharides (glycosaminoglycans, e.g. heparin) and proteins. Unpredicted cross reactions abound and every drug can be expected to interact with multiple proteins. This provides a note of caution to the use of any drug and encourages minimal exposure, since many unobserved and unanticipated side effects are occurring. These observations also question routine ingestion of herbal remedies, after all, plants use their secondary products as potent defenses against being eaten. Alkaloids disrupt nervous systems and cellular signaling. Plants are not naturally safe.